Assistive technology (AT) has tremendous potential to support increased independence in adults with severe cognitive impairments due to traumatic brain injury by compensating for these impairments. However, the AT literature cites lack of effective instruction as a barrier to successful, long-term AT use. Previous research has shown that a systematic instructional package (TEACH-M) was effective in training four individuals with severe cognitive impairments to learn a multi-step procedure using one form of assistive technology: adapted email. TEACH-M incorporates empirically-validated instructional techniques from research in special education and neuropsychological rehabilitation. Components include errorless learning, high rates of correct practice, spaced retrieval, regular assessment, and metacognitive strategy training. The goal of this study is to expand previous research by examining the effectiveness of the TEACH-M instructional package compared with conventional instruction in training 40 individuals with TBI to use a personal digital assistant (PDA). A randomized control, cross-over design with repeated measures will be used to compare performance across the two types of instruction. The participants will be randomly assigned to two groups. The training conditions (i.e., TEACH-M; conventional instruction) and PDA training tasks will be counterbalanced across these two groups to avoid order effects. Outcome measures will include pre-post training comparisons of performance efficiency (i.e., average percentage of correct steps completed per minute), 30-day maintenance probes, the number of sessions to reach criterion for task mastery per training condition, the proportion of participants who reach the criterion for mastery per training condition, and participant satisfaction ratings. The PIs hypothesize that the TEACH-M instructional package will result in more efficient task performance, better maintenance, and faster and higher rates of task mastery. Also, participants will prefer TEACH-M over conventional instruction. This study will address the critical lack of experimental evidence evaluating instruction as an essential component of successful AT use. Future research will test the generalization of training to use of AT devices in natural contexts. In addition, the proposed study will be replicated with different instructors, populations (e.g. individuals with early dementia), and types of AT (e.g., off-the-shelf "smart phones"; customized handheld organizers). [unreadable] [unreadable] The project will address the critical lack of research evaluating instruction as a component of successful AT use among individuals with cognitive disabilities due to TBI. The knowledge gained from this study will contribute to the body of scientific knowledge on effective instructional design for this population. [unreadable] [unreadable] [unreadable]